Current optical analysis instrumentation typically involves the use of an electronic detector for detecting the photon output at various wavelengths produced by an optical analysis instrument, such as a spectrometer or monochrometer. While these systems have reached an exceedingly high degree of sophistication and sensitivity, current practice does suffer from a number of limitations and inconveniences.
One class of devices analyzes light by breaking it into its constituent wavelengths or colors, and then measuring the amplitude of the light at the various wavelengths. Spectrometers perform this function by presenting at their output a spectrum of wavelengths where the position of the output light deviates from a reference wavelength by a distance which is a direct function of the wavelength.
One type of electronic detector which is used to quantitatively evaluate the amount of light at each wavelength is the charge coupled device (CCD) detector array. Generally, this device comprises an array of solid state device photodetectors which cover an area which corresponds to the area on which the spectrometer outputs light.
In order to reach a high degree of sensitivity and good signal-to-noise ratio, such electronic detectors are operated in an extremely cold environment which minimizes noise. More particularly, during operation, semiconductor junctions have "traps" which can tie up an electron or a hole, causing it to be unavailable for recombination. During operation, if an electron is caused to leave the valence band, a hole remains. If, however, the hole becomes trapped, it will not be available for electron recombination. Similarly, an electron trap in the conduction band will make an electron unavailable for recombination. The result of these two effects is to inhibit the conduction process in the device, increase response time and, in the case of electron traps, to reduce sensitivity.
A consideration in the design of such devices is the expected ambient temperature. Generally, temperature is the equivalent of kinetic energy in the various atoms of the material of the semiconductor detector. Increasing temperature results in imparting this kinetic energy to electrons, freeing them up to interfere with the detection process which depends upon the detection of free electrons created by photons of incident light, in order to measure the intensity of incident light. If the temperature of a detector is relatively high, and the intensity of light to be detected is relatively low, the electrons released by atoms due to the relatively high temperature create a thermal noise which effectively decreases the output signal-to-noise ratio of the device and thus reduce effective system sensitivity.
In the case where exposure times are very long, thermal noise tends to accumulate in the charge coupled device array, thus effectively masking the light input into the detector over what may be a period of minutes or even hours.
In a typical spectrographic application, the CCD detector is placed in a housing which maintains the detector in a vacuum. In addition, part of the housing forms a Dewar flask, which includes a compartment for liquid nitrogen which is typically at 77.degree. Kelvin. In this sort of application, the CCD detector used typically has an operating characteristic which is optimized at 1400 Kelvin. However, from a practical standpoint, 1400 Kelvin is not easy to achieve for extended time periods given the present state of spectrographic CCD housing system technology.
It is also desirable that the temperature of the detector be maintained at a constant level during operation. Prior art systems typically incorporate a heater element for heating up the CCD detector and keeping it from getting too cold. However, such heater elements are, as a practical matter, never used because of the inability of commercial systems to maintain the desired 1400 Kelvin temperature.